# What is Max Pooling? Definition & Technical Mechanics

> **Source:** [SPIDITS AI AI Glossary](https://spidits.com/ai-glossary)  
> **Category:** Neural Architectures  
> **Canonical Citation:** [https://spidits.com/ai-glossary/max-pooling](https://spidits.com/ai-glossary/max-pooling)

## Definition
Max Pooling is a sample-based discretization process in CNNs. It divides the input image into sub-regions and outputs the maximum value from each sub-region, reducing dimensional size.

## Architectural Mechanics & Deep Dive
Max pooling is a sample-based discretization process used in convolutional neural networks. It partitions the input image or feature map into a set of non-overlapping sub-regions, and for each sub-region, outputs the maximum value. This reduces the spatial dimensions of feature maps, lowering computational cost and providing translation invariance to small shifts.

## Industry Applications & Real-World Use Cases
CNN downsampling, feature size compression, and computer vision feature extraction.

## Frequently Asked Questions

### Why use max pooling in CNNs?
To reduce spatial dimensions (parameter counts) while retaining dominant visual features, helping build translation invariance.

### What is average pooling?
A variant that outputs the mathematical average of a sub-region instead of the maximum value.

## Related Terminology
- [cnn](https://spidits.com/ai-glossary/cnn)
- [computer-vision](https://spidits.com/ai-glossary/computer-vision)
- [dimensionality-reduction](https://spidits.com/ai-glossary/dimensionality-reduction)

---
*From the SPIDITS AI Knowledge Base (336+ terms indexed). Explore full technical definitions: [https://spidits.com/ai-glossary](https://spidits.com/ai-glossary)*